Published on

1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Firstly I’ll introduce peculiarities of SDM. They ‘re particularly interesting because the practice of geo-referencing them have caused a growing demand for powerful exploratory data analysis techniques overcomes classical statistical and data mining techniques and, among other things,support the analysis of socio economic phenomena by a spatial point of view. In this talk I’ll focus my attention on a specific task that is the discovery of spatial association rules For this purpose I’ll present ARES a system to extract association rules from census data and illustrate an application ARES to mine spatial association rules on North West England 1998 census data in order to study the mportality risk in Greater manchester county
  • What is IE. As a task it is… Starting with some text… and a empty data base with a defined ontology of fields and records, Use the information in the text to fill the database.
  • ML… although this is an area where ML has not yet trounced the hand-built systems. In some of the latest evaluations, hand-built shared 1 st place with a ML. Now many companies making a business from IE (from the Web): WasBang, Inxight, Intelliseek, ClearForest.
  • Data sparseness, robustness
  • CV i.e. it is divided into 5 folds (Four are used for training and one for testing in turn).
  • Initial ILP reasearch deals with concept learning in form of predicate definition learning
  • ATRE is a multiple-concept learning system, which solves the following problem:
  • Since the generation of a clause depends on the chosen seed, several seeds have to be chosen such that at least one seed per incomplete predicate definition is kept . Therefore, the search space is actually a forest of as many search-trees as the number of chosen seeds. The parallel exploration of the forest related to odd and even numbers. Spec. hierarchies are traversed top-dow. Search proceeds towards deeper and deeper levels of the specialization hierarchies until at least a user-defined number of consistent clauses is found. A supervisor task decides whether the search should carry on or not on the basis of the results returned by the concurrent tasks. When the search is stopped, the supervisor selects the “best” consistent clause according to the user’s preference criterion. This strategy has the advantage that simpler consistent clauses are found first, independently of the predicates to be learned. First learning step Consistent clauses in red
  • Second learning step
  • CV i.e. it is divided into 5 folds (Four are used for training and one for testing in turn).
  • If we guarantee the following two conditions: ……………………… then after a finite number of steps a theory T , which is complete and consistent, is built. If we denote by LHM( T i ) the least Herbrand model of a theory T i , the stepwise construction of theories entails that LHM( T i )  LHM( T i+1 ), for each i  {0, 1,  , n-1}, since the addition of a clause to a theory can only augment the LHM
  • In order to guarantee the first of the two conditions it is possible to proceed as follows. First, a positive example e + of a predicate p to be learned is selected, such that e + is not in LHM( T i ). The example e + is called seed . Then the space of definite clauses more general than e + is explored, looking for a clause C, if any, such that neg(LHM( T i  { C })) =  . In this way we guarantee that the second condition above holds as well. When found, C is added to T i giving T i+1 . If some positive examples are not included in LHM( T i+1 ) then a new seed is selected and the process is repeated. The second condition is more difficult to guarantee because of the non-monotonicity property. The approach followed in ATRE to remove inconsistency due to the addition of a clause to the theory consists of simple syntactic changes in the theory, which eventually creates new layers . The layering of a theory introduces a first variation of the classical separate-and-conquer strategy sketched above, since the addition of a locally consistent clause generated in the conquer stage is preceded by a global consistency check.
  • Learning multi-relational patterns from multi-relational data and background knowledge It allows to navigate the relational structure of data
  • Download

    1. 1. Mobile Broadband: The Global Evolution of UMTS/HSPA 3GPP Release 7 and Beyond ACRONYM LIST........................................................................................................................................... 3 PREFACE ..................................................................................................................................................... 5 1 INTRODUCTION........................................................................................................................................ 6 2 PROGRESS OF REL-99/REL-5/REL-6 UMTS ......................................................................................... 7 PROGRESS TIMELINE .................................................................................................................................. 7 3 THE GROWING DEMANDS FOR WIRELESS DATA APPLICATIONS .................................................. 9 WIRELESS DATA TRENDS AND FORECASTS .................................................................................................. 9 4 OVERVIEW OF 3GPP REL-7.................................................................................................................. 15 4.1 MIMO ON HSDPA.............................................................................................................................. 15 4.2 RAN ENHANCEMENTS ......................................................................................................................... 19 4.3 IMS/CORE NETWORK ENHANCEMENTS ................................................................................................ 24 5 EVOLUTION BEYOND REL-7 – SAE/HSPA+/LTE................................................................................ 26 5.1 ARCHITECTURE EVOLUTION (SAE) ...................................................................................................... 27 5.2 AIR-INTERFACE EVOLUTION OR LONG TERM EVOLUTION (LTE) ............................................................. 29 5.3 HSPA EVOLUTION (HSPA+) ............................................................................................................... 35 6 CONCLUSIONS....................................................................................................................................... 37 APPENDIX A: DETAILED VENDOR PROGRESS ON REL-99/REL-5/REL-6 ......................................... 39 APPENDIX B: GLOBAL UMTS OPERATOR STATUS ............................................................................ 48 APPENDIX C: GLOBAL EDGE-UMTS OPERATOR STATUS & DEVICES ............................................ 56 APPENDIX D: GLOBAL HSDPA OPERATOR STATUS & DEVICES ..................................................... 58 HSDPA DEPLOYMENTS ............................................................................................................................ 58 HSDPA DEVICES ..................................................................................................................................... 60 VENDORS ................................................................................................................................................. 61 ACKNOWLEDGMENTS ............................................................................................................................. 63 2
    2. 2. Acronym List 1xEV-DO 1x EVolution Data Optimized 1xEV-DV 1x EVolution Data Voice 3GPP 3rd Generation Partnership Project AGPS Assisted Global Positioning System AMR Adaptive Multi-Rate ARPU Average Revenue Per User BTS Base Transceiver Station CQI Channel Quality Indications CS Circuit Switched CSI Combination of Circuit Switched and Packet Switched services CSCF Call Session Control Function CTIA Cellular Telecommunication Industry Association D-TxAA Double Transmit Adaptive Array DCH Dedicated Channel E2E End to End E-DCH Enhanced Dedicated Channel (also known as HSUPA) E-DPCCH Enhanced Dedicated Physical Control CHannel E-DPDCH Enhanced Dedicated Physical Data CHannelCHannel EDGE Enhanced Data for Global Evolution ETSI European Telecommunication Standards Institute EUTRA Evolved Universal Terrestrial Radio Access EUTRAN Evolved Universal Terrestrial Radio Access Network FBC Flow Based Charging FBI Fixed Broadband access to IMS FDD Frequency Division Duplex GPRS General Packet Radio System GSM Global System for Mobile communications GUP Generic User Profile HARQ Hybrid Automatic Repeat Request HLR Home Location Register HS-SCCH High-Speed Shared Control Channel HSDPA High Speed Downlink Packet Access HSPA High Speed Packet Access (HSDPA + HSUPA) HSPA + High Speed Packet Access Plus (also known as HSPA Evolution) HSS Home Subscriber Server HSUPA High Speed Uplink Packet Access HTML Hyper-Text Markup Language IMS IP Multimedia Subsystem IP Internet Protocol ISIM IMS SIM ISP Internet Service Provider ISUP ISDN User Part ITU International Telecommunication Union J2ME Java 2 Micro Edition LCS LoCation Service LMMSE Least Minimum Mean Squared Error LTE Long Term Evolution (Air Interface Evolution) M2M Machine to Machine MAC Media Access Control MBMS Multimedia Broadcast/Multicast Service MITE IMS Multimedia Telephony Communication Enabler MRFP Multimedia Resource Function Processor MME Mobility Management Entity MMS Multimedia Messaging Service NGN Next Generation Network OMA Open Mobile Architecture OTA Over The Air PAR Peak to Average Ratio PARC Per-Antenna Rate Control 3
    3. 3. PCC Policy and Charging Convergence PCMCIA Personal Computer Manufacturers’ Card Interface Adapter PCS Personal Communication System PoC Push-to-talk over Cellular PLMN Public Land Mobile Network POTS Plain Old Telephone Service PS Packet Switched PSI Public Service Identities PSTN Public Switched Telephone Network QAM Quadrature Amplitude Modulation QPSK Quadrature Phase Shift Keying QoS Quality of Service RAB Radio Access Bearer RAT Radio Access Technology RNC Radio Network Controller RRC Radio Related Signaling SAE System Architecture Evolution SBLB Service Based Local Policy SDMA Spatial Division Multiple Access SIM Subscriber Identity Module SIP Session Initiated Protocol SMS Short Message Service SRNC Serving Radio Network Controller STTD Space-Time Transmit Diversity TFC Transport Format Combination TTI Transmission Time Interval UE User Equipment UMTS Universal Mobile Telecommunication System, also known as WCDMA UPE User Plane Entity USIM UMTS SIM UTRA Universal Terrestrial Radio Access UTRAN UMTS Terrestrial Radio Access Network VCC Voice Call Continuity VoIP Voice over Internet Protocol VPN Virtual Private Network WAP Wireless Application Protocol WCDMA Wideband Code Division Multiple Access WIM Wireless Internet Module WLAN Wireless Local Area Network 4
    4. 4. Preface The growing commercialization of Universal Mobile Telecommunications System (UMTS), also known as Wideband Code Division Multiple Access (WCDMA), has been the topic of an annual white paper by 3G Americas since 2003, when the focus was Third Generation Partnership Project (3GPP) Release '99. With both the rapid progress of the evolutionary 3GPP roadmap for UMTS to HSPA from Release 5 (2004 white paper) to Release 6 (2005 white paper) and now Release 7, and the commercial deployment of 107 UMTS/WCDMA networks worldwide with nearly 75 million customers at the writing of this paper, the need for semi-annual updates has become necessary in order to provide reasonably current information. In fact, on December 6, 2005, Cingular Wireless launched UMTS enhanced with High Speed Downlink Packet Access (HSDPA) in 16 major markets throughout the U.S., becoming the first operator in the world to launch this enhanced UMTS technology on a wide-scale basis. Already there are 41 operators offering HSDPA services in 31 countries of the world, with additional commitments from 62 more operators as of July 7, 2006 (see Appendix D). 3G Americas' first UMTS white paper, UMTS to Mobilize the Data World reported on the progress of UMTS: from its inception in 1995, to standardization by ETSI1 in January 1998, to the commercial launch by Japan's NTT DoCoMo and other operator trial launches. The paper provided documentation on the installation, testing and preparation of UMTS networks on several continents, and the prediction that UMTS and EDGE would serve as complementary technologies for GSM operators throughout the world. Figure 1. Global UMTS Subscriber Growth Forecast2 The rapid growth of UMTS led to a focus on its next significant evolutionary phase, namely, Release 5 (Rel-5). 3GPP Rel-5, initially deployed in 2005, has many important enhancements that are easy upgrades to the initially deployed Release 1999 (Rel-99) UMTS networks. Rel-5 provides wireless operators the improvements they need for offering customers higher-speed wireless data services with vastly improved spectral efficiencies through the HSDPA feature. It is expected that HSDPA Rel-5 will provide a 50 percent reduction in cost per megabit versus Rel-99, and HSDPA Rel-6 will further build upon reductions in the cost per megabit. In addition to HSDPA, Rel-5 introduces the IP Multimedia Subsystem (IMS) architecture that promises to greatly enhance the end-user experience for integrated multimedia applications and offer mobile operators a more efficient means for offering such services. 1 ETSI: European Telecommunications Standards Institute 2 Informa Telecoms and Media, World Cellular Information Service, June 2006 5
    5. 5. UMTS Rel-5 also introduces the IP UTRAN concept to realize transport network efficiencies and reduce transport network costs. The 3G Americas white paper titled The Evolution of UMTS – 3GPP Release 5 and Beyond was published in June 2004, updated in November 2004, and provided an overview and status update of the key 3GPP Rel-5 specifications and features discussed above. The Global Evolution of UMTS/HSDPA - 3GPP Release 6 and Beyond December 2005 white paper provided information on the commercialization and industry progress towards the evolution of UMTS to Release 6 (Rel-6) with discussion of future evolutions of the technology. Mobile Broadband: The Global Evolution of UMTS/HSPA Release 7 and Beyond takes the step forward to Release 7 (Rel-7) and the future beyond HSPA. In this new paper, we explore UMTS/HSDPA commercialization status and its continuing standards developments, focusing on Rel-7 and looking at what lies beyond with the Long Term Evolution (LTE) and System Architecture Evolution (SAE) initiatives. We also explore the growing demands for wireless data and successes already indicated for a variety of wireless data applications: the increasing Average Revenue Per User (ARPU) for wireless data services by operators worldwide, the cost per byte of UMTS data service, and technology benefits. The appendices include lists of both commitments and deployments for EDGE, UMTS and HSDPA, as well as the progress of leading UMTS vendors. UMTS evolution is ongoing with a clear roadmap well into the near future, and this white paper provides the guide for its readers to understand the evolutionary process. This paper has been prepared by a working group of 3G Americas' member companies. The material represents the combined efforts of many experts from the following companies: Andrew Corporation, Cingular Wireless, Ericsson, Gemalto, Hewlett-Packard, Lucent Technologies, Motorola, Nokia, Nortel, and Siemens. 1 Introduction Wireless data revenues are picking up and have now passed 10% of the total revenues for many operators. Today the most popular applications are text messaging (SMS), Web and WAP access, multi- media messaging (MMS) and content downloads, e.g., ring tones, music and video clips, with more than 50% of the mobile subscribers using these basic data services. With the introduction of UMTS, data revenues will play an increasingly important role for operators. Most UMTS operators today are offering some kind of mobile broadband service and several PC vendors offer laptops with built-in HSDPA capabilities that will boost data usage even further. Other services, with less penetration today but with high expectations for future growth, include mobile email access, mobile TV, mobile gaming and full track music downloads. While earlier 3GPP releases already provide efficient support for these services, 3GPP Rel-7 focuses on providing improved support and performance for real-time conversational and interactive services such as Push-to-talk Over Cellular, picture and video sharing, and Voice and Video over IP. These services are starting to become available as operators deploy IP Multimedia Subsystem (IMS) based on 3GPP Rel-5. The radio interface enhancements in 3GPP Rel-7 will provide improved capacity in terms of the number of packet data users that can be connected simultaneously as well as improved state transition delays in order to shorten initial delays for service establishment or re-activation. The introduction of Multiple Input Multiple Output (MIMO) will further enhance mobile broadband offerings by taking theoretical peak rates well above today’s 14 Mbps as well as improving the average cell throughput. 3GPP Rel-7 also provides new and enhanced Core Network and IMS features. IMS is extended to wireline subscribers to make it the common platform for both fixed and mobile networks going forward. IMS Multi-media Telephony defines a telephony service for Voice and Video over IP, allowing operators to provide telephony services in the packet switched (PS) domain that is consistent with the already existing services in the CS domain. Voice Call Continuity provides seamless mobility between the voice component of the IMS Multi-media Telephony and CS voice. CSI provides the means for a mobile station to combine a CS voice call with PS-based IMS services between the same two users, enabling picture and video sharing with CS voice calls in order to provide enriched voice services. Policy and Charging Convergence (PCC) allows operators to perform advanced dynamic QoS control and charging for packet data services. Looking further ahead, 3GPP is working on a new radio interface and new system architecture in order to cope with the rapid growth in IP data traffic and ensure competitiveness for the next ten years and 6
    6. 6. beyond. The evolution of the technology will bring theoretical peak rates to above 100 Mbps for downlink and 50 Mbps for uplink, and reduce latency to levels comparable with fixed broadband Internet, e.g., less than 5 ms in ideal conditions. In order to achieve this, an evolution or migration of the network architecture, as well as an evolution of the radio interface is studied in the System Architecture Evolution (SAE), Long Term Evolution (LTE) and HSPA Evolution (HSPA+) Study Items. 2 Progress of Rel-99/Rel-5/Rel-6 UMTS At the time of this writing, 3GPP Rel-99 through Rel-4 and Rel-5 UMTS are considered mature specifications, with Rel-99 initially standardized in early-mid 1999 and published by 3GPP in March 2000, Rel-4 completed in March 2001, and Rel-5 published in March 2002. The Rel-99 UMTS specifications provided an evolution path for the GSM, GPRS and EDGE technologies that enabled more spectrally efficient and better performing voice and data services through the introduction of a 5 MHz UMTS carrier. Commercial deployments of Rel-99 UMTS networks began several years ago and the number of commercially deployed UMTS systems has grown rapidly since then, as substantiated in the 107 commercial UMTS networks in the deployment status listing in Appendix B of this paper. Rel-4 introduced call and bearer separation in the Core Network, and Rel-5 introduced some significant enhancements to UMTS including HSDPA, IMS and IP UTRAN.3 Rel-6 was completed in March 2005 introducing further enhancements to UMTS including HSUPA (or E- DCH), MBMS and Advanced Receivers.4 The first commercial deployments of HSDPA occurred in Q4 2005, while initial deployments of IMS are already underway. There have also been numerous trials and demos of Rel-6 features such as HSUPA in 1Q-2Q 2006. Leading manufacturers worldwide support UMTS and to illustrate the rapid progress and growth of UMTS, detailed descriptions of recent accomplishments from each of the 3G Americas’ participating vendors on Rel-99, Rel-5 and Rel-6 UMTS are included in Appendix A of this white paper. As there are many technology milestones to report, a summary of some of the key technology milestones and developments is provided in this section. However, Appendix A has individual reports from leading vendors on their recent demonstrations, trials, product releases and contracts. Progress Timeline Figure 2. 3GPP UMTS Timeline5 3 3GPP Rel-5 and Beyond - The Evolution of UMTS, November 2004, 3G Americas, 4 The Global Evolution of UMTS/HSDPA - 3GPP Release 6 and Beyond, December 2005, 3G Americas, 5 3G Americas, June 2006 7
    7. 7. In November 2003, HSDPA was first demonstrated on a commercially available UMTS base station in Swindon, U.K. HSDPA was first commercially launched on a wide scale by Cingular Wireless in December 2005, followed closely thereafter by Manx Telecom and Telekom Austria. In June 2006, "Bitė Lietuva" of Lithuania became the first operator to launch HSDPA at 3.6 Mbps, a record speed. At this time, there are 41 commercial HSDPA networks with an additional 62 operators committed to deploying HSDPA. It is expected that almost all UMTS operators will deploy HSDPA. In early 2006, leading infrastructure vendors launched a new generation of base stations allowing for fewer sites, increasing capacity by 50-150 percent and optimized for cost efficiency at every site. Power consumption was cut further by 35-55 percent thereby enhancing power efficiency. These base stations allow GSM to be seamlessly upgraded to UMTS and operate in both GSM and UMTS networks in parallel. Frequency bands are currently supported in the 850, 1900 and 2100 MHz bands and will also support all coming frequency bands, including 900, 1700, 1800, 1700/2100 and 2500 MHz. The 700 MHz band will be introduced in time to support the rollouts when the 700 MHz band has been auctioned and cleared. One vendor cites the mobile-data throughput capability of the most cost-effective base station is more than 400 GB per day, resulting in a broadband radio network at a cost close to $1 per GB. With reportedly up to 70 percent lower base station site expenditures, the GSM/UMTS infrastructure costs are encouraging operators to deploy 3G UMTS technology today. Most vendors' HSDPA solutions require software-only upgrades, which can be downloaded remotely to the UMTS RNC and Node B. At 3GSM 2006 in Barcelona, infrastructure vendors demonstrated live HSDPA data downloading at a speed of 3.6mbps. HSDPA on-air tests have shown 950kbps with 80mph in a live UMTS network. Delivery of HSUPA is expected in 1Q2007, compliant to Rel-6. Handset manufacturers have already shipped the world’s first handsets that support HSDPA, and, in May of this year, the first commercial network with HSDPA handhelds was launched in South Korea. In addition to allowing data to be downloaded at up to 1.8 Mbps, the initial handsets offer such applications as satellite-transmitted Digital Multimedia Broadcasting (DMB) TV programs, have two to-three-megapixel cameras, Bluetooth, radios and stereo speakers. Many new devices are anticipated for launch in 2H 2006; Cingular has expectations to offer their customers several devices in mid-year. HSDPA data cards support different UMTS frequency bands, and can be used in both the U.S. and Europe. Some vendors’ wireless modules are expected to introduce more UMTS/HSDPA data capabilities in 2006 through enabling the machine-to-machine (M2M) market and further improving the diversity of possible applications. The world’s first HSDPA demonstrations with Multiple Input Multiple Output (MIMO) technology using a commercial radio base station were in April at CTIA Wireless 2006, and data rates were doubled from 10 to 20 Mbps. HSUPA technology was demonstrated live at several global trade shows in early 2006, with speeds peaking at more than 4Mbps, enabling the use of many more applications. Some vendors plan to have HSUPA for enhanced uplink data speeds available as early as the first quarter of 2007, in compliance with 3GPP Rel-6. Beyond HSPA, leading vendors are actively developing and testing IMS device implementation. The GSMA’s IMS (Videoshare) Interoperability Test Sessions yielded important early successes in demonstrating IMS functionality, as well as ensuring interoperable solutions that will increase the take-up of this next step in the GSM/UMTS evolution. Vendors are also supporting IMS development across multiple frequency bands to deliver valuable applications and services. Vendors have signed commercial IMS agreements throughout the world and are conducting hundreds of trials of various IMS network elements. IMS developer programs are available in Germany, USA, China and Singapore to encourage the creation of advanced IMS applications and services. Some vendors have announced plans to make their current infrastructure reusable for Long Term Evolution (LTE), for example, by introducing new modules that can be integrated into the existing base stations. These future network infrastructure architectures will find their way into the market around 2008 or 2009. Technology milestones and advances in the evolution of UMTS continue to develop as the number of 3G customers grows at a rapidly increasing rate. With the structure for services and applications beginning to grow more secure, the demand for wireless data services and other advance voice applications is showing a growth trend as well. Reference Appendix A for more detailed information on the progress of the GSM to HSPA evolution and beyond. 8
    8. 8. 3 The Growing Demands for Wireless Data Applications As 3G networks continue their rollout worldwide, manufacturers are enabling a slew of applications that are driving innovations in mobile handsets, and crossing barriers into a wide variety of vertical enterprise markets. Consumers are likewise driving the mobile content for entertainment, advertising, and MMS services. As these applications proliferate, voice is becoming a secondary factor when purchasing a mobile device, according to Naqi Jaffery, Telecom Trends International.6 High-speed wireless data services have elevated what Motorola calls “the device formerly known as the cell phone” to much more. Wireless-phone makers offer a new generation of handsets that combine a variety of functions in a single device. Handsets can send email, store music on removable memory, store video clips, check satellite positioning and even monitor a user’s stress level. A wireless device is now a must-have device for lifestyle–conscious consumers as well as world travelers and business people. Wireless Data Trends and Forecasts “Each year, more and more consumers are experiencing the incredible benefits that only wireless can offer,” said CTIA President, Steve Largent. “The mobile communications revolution is in full swing, and now nearly 70% of America is taking part in it.” According to a CTIA survey, revenues from wireless data services jumped more than 86% in the past year, amounting to more than $8.58 billion in 2005, up from $4.6 billion in 2004. American customers sent nearly 50 billion text messages in the six month period ending December 2005, and consumed nearly 1.5 trillion wireless minutes in the year 2005. These two measurements represented a 97% and a 36% increase respectively. With the U.S. adding more than 25 million new subscribers in 2005, the U.S. wireless industry had a stellar 2005, passing the 200 million subscriber mark and 70% market penetration thresholds, receiving positive market response to both PC cards and the first handset-based 3G applications, and with providers reporting they had crossed the 10% data ARPU level. “Total voice service revenue declines late in the forecast period [U.S. wireless industry in 2008-2009] will jolt an industry accustomed to 25 years of voice revenue growth and further emphasize the importance of data services to the future of the industry,” observed Scott Ellison, IDC.7 Wireless Data Revenue U.S. data revenue is expected to grow 45% to $12.6 billion in 2006 from the previous year, according to research firm Ovum.8 In comparison, Ovum expects total wireless revenue to grow 10% to $135.86 billion in 2006. On a global basis, Kagan Research forecasts total wireless data revenues will be the wireless industry’s quickest growth area, increasing from $8.4 billion in 2005 to $46.6 billion by 20149. That would account for about 24 percent of a total $191 billion in 2014 wireless service revenues, against just 7 percent in 2005. Kagan also suggests that interactive applications for games, t-commerce and advertising will become an increasing portion of subscriber-related multi-channel revenue at 3.2 percent by 2015. Including video on demand and pay per view revenues will further increase the sector to nearly 10 percent in the ten year outlook. In November 2005, Yankee Group launched the Global Wireless/Mobile Premium Forecast, predicting that data services would comprise 21% of total worldwide wireless operator service revenue of US$698 billion by 2009. Text messaging revenue will top US$36 billion in 2009, ringtone revenue to carriers and content providers will reach almost US$28 billion.10 Carriers are beginning to reap the benefits of faster networks and “devices formerly known as the cell phone”. Cingular Wireless, the largest carrier in the U.S., added a record 1.8 million net subscribers in Q4 2005. Driven by its best ever overall churn performance and record gross additions, the company ended the year 2005 with 54.1 million customers. Cingular reported that ARPU reflects continued pressure on voice revenues as the wireless market becomes more penetrated and lower-revenue 6 “The Future of Mobile Handsets,” Telecom Trends International report, May 1, 2006 7 IDC Forecast Anticipates Decline in Total Voice Revenue in 2008-2009 for the U.S. Consumer Wireless Industry, IDC press release, March 29, 2006 8 3G Retains its Buzz but Potential Remains Unclear, Dow Jones NY, January 27, 2006 9 Media Trends 2006, Kagan Research, April 2006 10 Global Wireless/Mobile Premium Forecast, Yankee Group report, November 2005, © Copyright 1997-2006. Yankee Group Research, Inc. All rights reserved. 9
    9. 9. customers enter the category, though these impacts were substantially offset by continued increase in data ARPU. The company’s data ARPU was up 63% over the year-ago fourth quarter at $4.71 and represented 10.4% of overall revenue. For T-Mobile USA, data revenues now represent 9.6% of postpaid ARPU, or $5.21 per customer, compared to 8.8% in the third quarter of 2005 and 6.6% in the fourth quarter of 2004. Central to the growth in data services revenue was a net increase in postpaid converged device users of more than 123,000 during 4Q 2005. In total, T-Mobile USA had 1.1 million converged device users at the end of 2005 including both the Blackberry and T-Mobile Sidekick devices. Rogers Wireless in Canada reported in 1Q 2006 that wireless postpaid voice and data subscriber ARPU increased in Q1 2006 to $62.20, an increase of 5.1% over first quarter 2005. Data revenues increased 72% and roaming revenues increased 33.1% in the first quarter year over year. Data revenue totaled $98.5 million for the first three months of 2006 or approximately 10.3% of Rogers total wireless network revenue compared to 7.1% in the same period of 2005. Rogers stated this increase reflects the continued rapid growth of Blackberry, text and multimedia messaging services, wireless Internet access, downloadable ring tones, music and games, and other wireless data services and applications. Telcel, the América Móvil operation that is market leader in Mexico, reports year end 2005 data ARPU at 11% of total company revenues, up from 9% in 2004, with a “vision of a steady pace of growth as new applications and services (i.e. mobile payments) are launched and the usage of GPRS/EDGE networks elevates”. Informa Telecoms & Media’s Data Metrics offers a look at the 20 global operators leading the market with data as the largest percentage of total revenue. As of March 2006, Filipino operators Smart and Globe continue to report the highest percentage of revenue for data on a global bases, 50% and 38% of revenue from data respectively in 4Q 2005. As of March 2005, Globe commercially launched HSDPA, enabling them to reduce their cost per bit for data services. Figure 3. Global Data Percentage of Revenue11 Some analysts are predicting that mobile data revenues worldwide will represent up to 35% of total monthly ARPU by 2010. It is no surprise that NTT DoCoMo is among the leaders, having done the 11 Informa Telecoms and Media, World Cellular Information Service, June 2006 10
    10. 10. groundbreaking work in this space since 1998. NTT DoCoMo earns more than $1 billion per month from data services with a stated goal to derive 80% of their revenues from data services by 2010. 3G Devices Gartner predicts that sales of HSDPA handsets will reach 2.1 million this year and will jump to 89.3 million by 2009.12 In a bid to encourage mobile customers to show more interest in multimedia services and applications, two device vendors have taken a more pro-active stance. Motorola has a flash new mobile music and video messaging application (which uses Flash) and Nokia has launched an interactive online community for mobile movies. Motorola’s StudioMOTO is a free online music studio that lets users create their own multimedia for their mobile and Nokia’s Nseries Studio enables filmmaking with mobile devices. In addition to handsets, soaring notebook PC shipments reflect strong demand for mobility on the computer market and generate rapid growth in the device population, claimed a Berg Insight report in January 2006. According to the study, HSDPA is the best suited wireless broadband technology for most regular users. T-Mobile (Germany) already began shipping HSDPA-ready PC data cards in late 2005, as did Cingular Wireless (US). Tobias Ryberg of Berg Insight said, “One of the key issues for the telecom industry is how [notebook PCs] should be connected to the Internet. There is much talk about various emerging technologies, but HSDPA is actually here right now and combines high performance with good coverage. No other technology can be expected to achieve the same footprint in the near future.” Ryberg said that European mobile operators were aiming to connect 50 million notebook PCs to the Internet via HSDPA. According to the report there are currently more than one million 3G data card users worldwide. Virtually all 3G network operators in Europe offer mobile broadband services. Beginning in the first half of this year, PC vendors including HP, Dell, Lenovo and Fujitsu will offer notebook computers with optional integrated HSDPA wireless broadband capability, which will boost current 3G download speeds by approximately four times. This was further supported by ABI Research which reported that the spread of high speed mobile data services is driving the increased adoption of wireless modems in laptop computers. The original wireless modems for laptops were add-ons in the shape of PC cards, and according to Phil Solis, senior analyst at ABI Research, “there are still several good years left in the PC card market.” Now, progressively more wireless modems are being built right into the computer, and it is there that the long term opportunity lies. This will produce a change in the dynamics of the market. ABI Research estimates that shipments of embedded modems will equal those of PC cards by 2009. Internet access via the mobile phone threatens to overtake wireless access from a notebook PC, according to the annual Face of the Web study of Internet trends from Ipsos Insight. The massive install base of mobile phones throughout the world is driving mobile access at a phenomenal rate.13 “In cellular’s early years, the primary focus was the cellular modem, and on improving its design,” according to Alan Varghese of ABI Research. “In later years, as cellular networks switched on high- speed data, the design focus veered away from the modem and towards applications such as digital imaging, music and video. Thus, it is no longer an either/or game: the chipset design focus has to be on both the modem and the applications in order to enable robust wireless connections, low power consumption and enjoyable user-experiences. 3G chipset architects find they have a full plate on their hands.”14 3G Applications Telephia reports that mobile data usage in the U.S., such as text and multimedia messaging, mobile Web, and downloads reached the 50% adoption mark in Q4 2005, rising seven percentage points since the beginning of the year. According to the latest data from Telephia’s Customer Value Metrics report, SMS activity leads the way for all mobile data usage with 41% of wireless subscribers using text messaging on their cell phones at the end of 2005. During Q4 2005, 22% of all cell phone users paid for accessing the Web via cell phone, 13% used MMS services (which raised 5 percentage points since Q1 2005), and 11% downloaded content from their cell phones (up 3 percentage points from the beginning of the year).15 The global market for mobile-phone premium content, including music, gaming and video, is expected to 12 Forecast: Mobile Terminals, Worldwide, 2000-2009 (4Q05 Update) Gartner, January 12, 2006 13 Mobile Phones to Rival PC for Internet Access, iTWire, Stan Beer, April 20, 2006 14 3G Chipset Design for Music, Gaming, Video and TV: No Longer an Either/Or Game, ABI Research press release, February 23, 2006 15 American Mobile Data Usage Surges, Telephia, April 7, 2006, 11
    11. 11. expand to more than $43 billion by 2010, rising at a compound annual growth rate of 42.5 percent from $5.2 billion in 2004, iSuppli Corp. predicts.16 Downloading music to a mobile handset, accessing and editing e-mails on a mobile handset, watching television or the person at the other end on the display during a phone call -- which services will wireless customers be using in the future, and what will they be willing to pay for them? To find the answers, Siemens surveyed over 5,300 mobile communication subscribers in eight countries about innovative wireless applications and their expectations with respect to the content and functionality of these applications.17 A few trends are clear; mobile television and e-mail access on a mobile handset number among the most popular applications. The survey on "innovative wireless services" was conducted in Brazil, Canada, China, Germany, Italy, Korea, Russia and the United States. It was found that consumers across all countries and continents are keenly interested in mobile e-mail access -- on average, 74 percent of the surveyed wireless users want to be able to send, receive and edit e-mails on their wireless devices. Users in North America and Europe, in particular, have high expectations with respect to mobile e-mail access. Although the focus is on using this technology in connection with their work, many wireless subscribers would also consider personal usage. In many countries, Mobile TV is one of the most attractive applications - with an average of 59% of all respondents expressing interest in utilizing this application. This application is especially interesting to wireless providers due to the strong willingness on the part of users to pay for it. In Korea, where Mobile TV is already being offered, more than 90% of all respondents voiced their interest in mobile television. The study shows that the ability to download music tracks to mobile handsets also offers high potential with 62% of all respondents indicating they would download music files to their wireless devices. In all of the countries in which the survey was conducted, mobile music downloads numbered among the three most popular applications. Other applications that rank high-- particularly among the young generation-- include Mobile Gaming as well as the IMS-based services Group Communication, Enriched Voice Calling and File Sharing. A new report by Datamonitor expects the number of mobilized email accounts to explode over the next three years. According to the report there are roughly 650 million corporate email inboxes worldwide today. Based on the assumption that at least 35-40% of these inboxes could potentially be mobilized, Datamonitor estimates the global addressable market for enterprise mobile email at around 260 million subscriptions. According to Datamonitor, mobile operators are in a position to make the most of the 18 upsurge in growth anticipated for mobile email. Motorola is developing a chip that would allow users to pay at cash registers by swiping their cell phones. Such payments already are prevalent in Asia, where customers can buy groceries, movie tickets and train fare with their mobile phones.19 With people spending billions on ringtones, wallpaper and games for their phones, analysts and retail executives say they believe it will not be much of a leap to get them to use their phones to buy shoes, books and laptops. “This will show up on the radar screen in 2006,” said Roger Entner, Ovum. “The more different pieces we add to these Swiss Army phones, the easier it is to get user acceptance for the next application. And especially around next Christmas, the convenience of shopping on a computer or a cell phone will beat the mall hands down.” 20 Trip Hawkins, PC-gamer pioneer and founder and CEO of start-up Digital Chocolate is helping to shape the nascent mobile gaming industry.21 Hawkins considers mobile gaming a community experience—the mobile phone is a social computer. What started on the Internet with dating services, chat and IM has spread like wildfire to the mobile phone with messaging, mobile email, and personalization—all about social identity. “The mobile phone is in a position to be the dominant platform in the same way the PC turned out to be the dominant platform of desktop computing. What the mobile phone has going for it as the social computer is ubiquity, because everyone has one, and because it supports a mobile lifestyle,” states Hawkins. “On the regular Internet, people are used to getting everything for free. But with the mobile phone, people are willing to pay for services. So there’s a huge business opportunity.” Hawkins sees the next important change as faster networks as there will be a quantum leap as we go from 2G to 3G networks meaning the gaming experience will get a lot better. 16 iSuppli Corp, March 8, 2006, 17 End-User Requirements & Expectations, presentation, Siemens Network Evolution Forum, 3GSM World Congress 2006, Barcelona 18 Mobile Email on the Verge of Mass Market Adoption, Datamonitor press release, February 2, 2006, 19 Mobile Phones the Next Wave in Contactless Payments, CTIA Smartbrief, February 8, 2006 20 “Those Born to Shop Can Now Use Cellphones,” Bob Tedeschi, New York Times, January 2, 2006 21 Get Your Digits in Shape: Mobile Games are Hot and Getting Hotter, Business Week, Business News Online Newsmaker Q&A, Steve Hamm, April 3, 2006 12
    12. 12. UK- based Juniper Research predicts that worldwide revenues for mobile gaming, which is composed of casino-style games, sports betting and lotteries will reach $19.3 billion by 2009. Juniper expects mobile gaming to make up roughly a third of the entire estimated $60 billion mobile entertainment market for that year. Mobile gaming already hit $468 million in 2004, and is projected at $2.07 billion at the end of 2005.22 Telephia found the mobile gaming industry is “soaring” with wireless customers buying more than 8.2 million mobile games in March 2006, up 53% from 5.4 million in January 2006 according to their report. Unique buyers also “surged” from an estimated 3.5 million in January to 5 million in March 2006 according to Telephia’s report. 23 In August 2005, the Pelorus Group predicted that mobile Web connectivity and services would grow from $1 billion to $15.3 billion in five years time (2005-2010).24 Social networking websites such as, which will soon go mobile, and others like Flickr for photo-sharing, for college students, or Rabble which lets users create profiles so they can share photos, videos and blogs with other members of the Rabble group, could become key applications drivers for new 3G wireless networks. Cell phones are the ideal tools for social networking and building online communities, because not only are people rarely without their phones, but today’s handsets also come equipped with sophisticated features such as cameras, digital music and video players and recorders that can be used for documenting one's life. Mobile-handset makers Nokia and Sony Ericsson are even embedding technology into some of their phones to make it easier for users to upload pictures and text to blogs. This makes the mobile phone a great tool for users to create their own content. The market for full track music downloads to mobile devices was twenty times larger at the end of 2005 than it was twelve months earlier noted ABI Research in its “Mobile Music Services” survey of world markets. Global revenues from over-the-air downloaded full track songs last year were $251 million up from $12.4 million in 2004.25 Nearly a quarter of Americans have downloaded a ringtone to their mobile phone, according to Ipsos Insight in TEMPO, their quarterly study of digital music behaviors since 2002 published in August 2005. This marked a dramatic increase from the 5 percent who had downloaded music one year earlier (in 2004).26 Mobile music now accounts for approximately 40% of record company digital revenues (which totaled $1.1 billion in 2005, up from $380 billion in 2004). Record companies are seeing sharply increased sales of master ringtones which account for the bulk of their $400 million plus mobile music revenues.27 Adoption of mobile video on phones has been somewhat low to date with only 2% of users claiming a subscription. JupiterResearch has found that 41% of mobile phone users are interested in some form of video service on their mobile phones. JupiterResearch expects the growing demand for video to generate $501 million in revenues by 2010 from $62 million in 2005. Among mobile subscribers today, 17% were interested in watching ‘live TV’ on their phones while 11% indicated interest in short video clips.28 Mobile advertising had $45 million in US ad revenue in 2005, but this is expected to grow to $1.26 billion in three years, according to Ovum’s forecast, and then carry on growing.29 Mobile ad budgets for Third Screen company had risen from an average of $20,000 for a campaign a year ago to ranges of $150,000 to $250,000 today. In January 2006, Third Screen signed a $1.6 million contract for a one-year campaign. One reason for the growing interest in cell phone ads is the relatively high click-through rate -- around 4 percent for mobile phones compared to 1 percent on the internet.30 Another tremendous area of growth for wireless data will be in machine-to-machine (M2M) wireless mobile connections. In Europe alone, Berg Insight predicts at least 25 million machines with connectivity to mobile networks by 2009.31 This number is currently estimated at 5 million. Although the use of these 22 Interactive Games Announces Launch of Cell Phone Gaming Division and Strategic Alliance with Via-Cell, Market Wire, April 28, 2005 23 Metric: Mobile gaming purchases up 53%, Fierce Wireless, March 8, 2006 24 Mobile Web Market to Hit $15.3B, Fierce Wireless, August 31, 2005 25 Ringing in the Changes: The Explosive Growth of Mobile Music Downloads, ABI Research press release, March 28, 2006 26 More Americans Download Music, Fierce Wireless, August 10, 2005 27 Mobile Music Downloads Soar, Cellular News, January 24, 2006 28 JupiterResearch Predicts the Mobile Video Market will Generate Revenues of Over $500 Million by 2010 as a Result of Expanded Video Offerings on Mobile Phones, JupiterResearch press release, March 28, 2006 29 “We’ll Be Returning You to our Mobile Phone Call-Right After These Important Messages,” Telecom TV, Martyn Warwick, January 18, 2006 30 “Marketers Want to Appear on Small Screen” New York Times, Matt Richtel January 16, 2006 31 Mobile M2M connections to hit 25 million by 2009, CNET Networks, Jo Best, May 11, 2006 13
    13. 13. nodes is largely for GSM/GPS tracking technology, other growth areas will be in the security and retail sectors. This will include use of 3G applications and services. Robin Duke-Woolley of Harbor Research predicts that revenues from M2M networking devices will grow 27% annually, and could reach $10.6 billion worldwide by 2011.32 IP Multimedia Subsystem (IMS) It has been estimated that by the end of the decade, more than 400 million people will regularly use SIP- based services across IP multimedia networks. IMS Research conducted an assessment of the trends affecting the uptake, including the timeline for IMS deployment. The assessment concluded that after initially slow uptake, the most popular service will be ‘push to talk’ followed by picture and video sharing, also known as ‘push to’ service.33 Services included in the assessment were conferencing, instant messaging, interactive gaming, location services and mapping, mo-blogging and voice over WLAN among others. Total revenues for these services on cellular networks enabled with IP multimedia subsystems were estimated by IMS Research to ramp up to exceeding $50 billion by 2010. Report author John Devlin said, “The biggest single factor affecting the uptake of services on cellular networks will be operator strategy with regard to the positioning and pricing of these services and also the upgrade path to 3G and beyond. IMS is seen by some operators in the industry as an enhancement for 3G networks and they are intending to use it as an enabler to encourage subscriber migration and service uptake.” Figure 4. Worldwide Uptake of IMS Services In Terms of Subscriber Numbers34 The IMS market is expected to peak in 2010 when investments in IMS equipment are expected to reach $4.5 billion, according to Informa Telecoms & Media.35 Informa predicts that successful services running on IMS will include VoIP and IPTV in the fixed environment and Push-to-talk over Cellular (PoC) and Instant Messaging in the mobile market. In the fixed market, Informa predicts that 39 million users will be enjoying IMS enabled services by 2011 compared with the 188 million users in the mobile sector. The next two years will be pivotal to the adoption of IMS worldwide. As of early April 2006, there were more than 50 contracts and 100 ongoing trials, and operators are expected to form strategies within the next two years. Operators are expected to enjoy cost savings from the layered architecture that IMS offers while maintaining a future-proof network. In the mobile domain, Informa expects that PoC will be a major 32 Harbor Research M2M Forecast, 2006 33 “Push To” Services take the Lead, IMS Research press release, March 16, 2006 34 IMS Research, Worldwide Uptake of Wireless Services, Wm. Morelli, May 2006 35 IMS Opportunities and Challenges, Informa Telecoms & Media report, March 2006 14
    14. 14. driver for IMS adoption while video sharing is already operating over IMS networks in a handful of cases around the world. VoIP over Cellular Industry interest in ‘VoIP over cellular’ is increasing. Reasons include the prospects of higher ARPU through richer communication (evolution currently driven by Internet players); lower OPEX through the offering of all mobile services from a common PS platform; and fixed/mobile convergence. The movement is to standardize an ‘IMS Multimedia Telephony’ service in 3GPP for many reasons: standardized services have benefits over proprietary solutions in terms of mass market potential; IMS is the standardized IP service engine for 3GPP access; and the service should make use of IP’s multimedia capability and flexibility, while retaining key telephony characteristics. 3GPP is the body with major mobile telephony expertise to accomplish this standardization process. The market indicators described in this section are representative of anticipated growth in the demand for wireless data services. The 3G networks and devices, applications and business strategies for IMS and UMA are in place or in development and researchers agree that the industry is poised for an explosion in the area of wireless data. Recent industry progress of UMTS from Rel-5 towards Rel-6 clearly supports the market trend claims reported in this section. 4 Overview of 3GPP Rel-7 As demonstrated in the previous 3G Americas’ white papers on UMTS/HSDPA [Ref 3, Ref 4], Rel-5 and Rel-6 introduced several features that provided significantly larger capacity and better performance compared to Rel-99. Rel-7 continues to build on the strong foundation of Rel-5/Rel-6 by introducing further capacity enhancing features such as MIMO for HSDPA. Moreover, Rel-7 focuses on RAN enhancements such as Continuous Connectivity, Gaming over IP (GoIP) and Delay Optimizations that enhance the capabilities for providing efficient and optimized real-time services such as VoIP, gaming and push-to-talk (PTT). Finally, Rel-7 also introduces IMS/Core Network enhancements related to multi- media telephony, combining of circuit and packet switched services (CSI), policy and charging and voice call continuity. It is proposed to keep the current Radio Network and Core Network separation to enable fast rollouts and easy implementation to terminals. This section provides an overview of the Rel-7 features discussed above. 4.1 MIMO on HSDPA The term “MIMO” is an acronym for multiple-input, multiple-output, and it is used to refer to any wireless system with multiple antennas at the transmitter and receiver. At the transmitter, multiple antennas can be used to mitigate the effects of fading via transmit diversity and to increase throughput via spatial division multiple access (SDMA). Beamforming and sectorization are two examples of SDMA. Beamforming can also be used for range extension. At the receiver, multiple antennas can be used for receiver combining which provides diversity and combining gains. If multiple antennas are available at both the transmitter and receiver, then different data streams can be transmitted from each antenna, with each stream carrying different information but using the same frequency resources. This technique, known as spatial multiplexing (SM), can potentially increase a user's peak data rate compared to conventional single- stream transmission. In the context of HSDPA Rel-7, MIMO specifically refers to SM. In this section, we first describe basic principles of spatial multiplexing followed by a discussion of the two main MIMO proposals that were investigated for Rel-7. We then describe how MIMO is used in a system context and the network environments that are best-suited for MIMO. We assume throughout that the MIMO mode uses two Node B antennas and two antennas per user equipment (UE). HSDPA uses rate adaptation based on a user’s channel quality information (CQI), which is fed back from the UE to the Node B. In conventional single-antenna HSDPA, the highest rate achievable under realistic conditions employs 16QAM and coding rate 1/2. Using a spreading period of 16 chips per symbol and 15 parallel codes for the HSDCH, this yields 7.2 Mbps (3.84Mchips per sec * 15 codes / 16 chips per code * 4 bits per symbol * 1/2 information bit per coded bit). Spatial multiplexing achieves higher data rates by reusing spectral resources over multiple spatial dimensions. For example, using two transmit antennas, one could transmit two separate data streams, each with data rate 7.2 Mbps resulting in a peak rate of 14.4Mbps. Because the two data streams are modulated in the same bandwidth using the same set of spreading codes, MIMO for HSDPA is sometimes called “code reuse”. At the receiver, multiple antennas are required to demodulate the data streams based on their spatial characteristics. In general, the 15
    15. 15. minimum number of receiver antennas required is equal to the number of separate data streams. Different receiver techniques can be used, but they typically employ a minimum mean-squared error (MMSE) linear processor followed by a non-linear interference canceller. The two specific MIMO proposals for Rel-7 are Per-Antenna Rate Control (PARC)[R1-010879] and [R1- 060708] and Double Transmit Adaptive Array (D-TxAA) [R1-030556]. Their baseband block diagrams are shown in Figures 5 and 6, respectively. Each MIMO proposal consists of a single-stream mode and a SM mode. For PARC, the single-stream mode uses space-time transmit diversity (STTD) block coding following the channel coding, interleaving, mapping to symbols, and spreading (C.I.M.S.) of the information bits. In the SM mode, the information bits are first demultiplexed, prior to coding. Because the rate for each antenna can be adapted independently, the number of bits sent to each antenna can be different. For D-TxAA, the single-stream mode uses Rel-99 TxAA to achieve closed-loop transmit diversity. The signals for each antenna are weighted by a complex amplitude chosen to best match the instantaneous channel characteristics. The pairs of amplitudes are chosen from a pre-defined set. Spatial multiplexing for D-TxAA transmits two simultaneous streams whose weight vectors are mutually orthogonal. Simulation results from various companies have shown similar performance for the PARC and D-TxAA schemes. In Rel-7, MIMO based on PARC has been adopted for UTRA TDD mode and Dual-codeword MIMO based on D-TxAA (with weights being signaled on the HS-SCCH in the downlink) has been adopted for UTRA FDD mode. In PARC, channel quality information is fed back from the UE to the NodeB to indicate which transmission mode to use and what data rate(s) can be supported. D-TxAA requires additional feedback to indicate which weighting vector(s) to use. D-TxAA could potentially achieve slightly higher throughput than PARC for low mobility channels. info info bits C.I.M.S. bits C.I.M.S. STTD Single-stream TX Single-stream TX SM TX SM TX C.I.M.S. C.I.M.S. + 1:2 1:2 Demux Demux info C.I.M.S. info C.I.M.S. + bits bits Figure 5. PARC36 Figure 6. D-TxAA37 To demonstrate performance of these MIMO techniques, Figure 7 and Figure 8 depict the average cell throughput and CDF of user throughput respectively for a 2x2 D-TxAA system compared to a reference 1x2 Linear Minimum Mean Square Error (LMMSE) system when using a space-time channel model (SCM). The scenario used in the evaluation is a micro cellular network scenario according to the model for the “Urban Micro” environment. The inter-site distance was 1000 meters without additional penetration loss. It can be observed that 2x2 D-TxAA MIMO provides some gain (approximately 10%) over the reference 1x2 LMMSE system in average cell throughput and increases the peak throughputs of the individual users. The results are expected to be comparable to PARC. 36 TSG-R1(01)0879, 3GPP TSG RAN WG1, Turin, Italy, Aug 27-31, 2001 37 TSGR1#32(03)0556, 3GPP TSG RAN WG1, Marne la Vallee, France, May 19-22, 2003 16
    16. 16. Figure 7. Average cell throughput in SCM Urban Micro, RR scheduler38 Figure 8. CDF of user throughput39 It should be noted that the spatial multiplexing techniques discussed above could also be used for lower data rates by reducing the coding rate and using smaller data constellations. For example, one could 38 Nokia, " WCDMA MIMO performance", March 27-31, 2006, Athens, Greece, 3GPP TSG RAN WG1 Meeting #44bis, Tdoc R1-060799 39 3GPP RAN WG1 #45, R1-061451, Lucent Technologies 17
    17. 17. transmit two data streams, each with QPSK and coding rate ¼. The same data rate could be achieved with a single data stream using QPSK and coding rate ½. For lower data rates, it is more efficient to transmit using a single stream rather than with spatial multiplexing. In other words, for a given low rate and a given total transmit power, single stream transmission achieves a lower frame error rate. Therefore MIMO for HSDPA uses single-stream transmission for lower data rates and spatial multiplexing for only the higher data rates. Because multiple antennas are available at the Node B, transmit diversity can be used for single-stream transmission. The “crossover point” at which it becomes more efficient to transmit with spatial multiplexing depends on many factors including the number of UE antennas. For dual-antenna UEs, the crossover point is typically about 5dB SINR (geometry), corresponding to an achievable data rate of about 4Mbps. In conventional cellular networks with fully loaded 3-sector cells and universal frequency reuse, the geometry is greater than 4dB in about 35% of the locations. Therefore SM is used about 35% of the time, and the fraction of users that achieve a peak rate equal or higher than the single-stream peak rate of 7.2 Mbps is about 25%. This is compared to the 15% that can achieve 7.2 Mbps in a conventional baseline HSDPA system with dual-antenna UEs. In terms of total average cell throughput, MIMO provides about a 33% improvement over this baseline. If cell throughput is a more important performance metric than peak data rate, it is much more efficient to use sectorization instead of SM. Using 6-sector cells each with a single antenna, the throughput improvement over the baseline is about 80%, but the tradeoff is that the peak rate remains the same since only a single antenna is used in each sector. Because the gains of MIMO are relatively modest for conventional networks with fully loaded cells, Rel-7 MIMO studies are now done in the context of "contained" environments where the intercell interference is limited. Examples of these environments include malls, academic campuses, and airports. In these environments, MIMO performance gains are improved, in terms of the fraction of time SM is used and overall throughput. Figure 9 shows a system-level comparison between conventional single-antenna and MIMO systems for the case of isolated cells and a conventional interference-limited environment39. The MIMO system used in these simulations is a PARC system (described below); however, similar results are expected for a D- TxAA system. Note that the peak rate and cell edge rates are defined as the 90% and 10% points of the cumulative distribution of achievable rates. Figure 9. MIMO system-level performance40 Operators Perspective on MIMO Wireless network operators see a need for MIMO (Multiple-Input Multiple-Output) due to its many advantages. Wireless systems using MIMO represent an economical way to increase user capacity, range and throughput in a variety of environments, most notably those which are enclosed and having low radio interference such as small and/or isolated cells. MIMO has been variously defined as “two or more unique radio signals, in the same radio channel, where each signal carries different digital 40 3GPP RAN WG1 #45, R1-061451, Lucent Technologies 18
    18. 18. information“ and “two or more radio signals which use beam-forming, receive combining, and spatial multiplexing”. The use of multiple antennas at both transmitter and receiver allows: • Multiplicative increase in peak data rate • Significantly higher spectrum efficiency, especially in low-interference environments • Increased system capacity (number of users) In UMTS systems, operators see a great need for MIMO in “contained environments” such as: • Hot spots similar to those serviced by today’s WiFi systems (airports, hotel lobbies, etc) • Academic campuses, in various self-contained areas (quads, auditoriums, cafeterias, etc) • Stadiums and arenas, again, which offer self-contained environments • Malls and shopping areas, favored by large numbers of younger, internet-savvy users • Mass transportation (trains, etc) with users looking for interaction and entertainment • Enclosed parks and recreation areas • Residential homes, supplanting DSL/Cable services Operators believe that, notwithstanding the basic differences in the physical layers used by UMTS and LTE, the benefits envisioned from MIMO in LTE, can also be obtained from MIMO in UMTS systems, starting in Rel-7. 4.2 RAN Enhancements Continuous Connectivity for Packet Data Users Packet-oriented features like HSDPA and HSUPA (HSPA) in UMTS systems provide high data rates for both downlink and uplink. This will promote the subscribers’ desire for continuous connectivity, where the user stays connected over a long time span with only occasional active periods of data transmission, and avoiding frequent connection termination and re-establishment with its inherent overhead and delay. This is the perceived mode to which a subscriber is accustomed in fixed broadband networks (e.g., DSL) and may make a significant difference to the user experience. The Fractional-DPCH feature was introduced in Rel-6 to support a high number of HSDPA users in the code limited downlink, where effectively a user in the active state, not being transmitted with any data, is consuming only a very small portion of the downlink capacity. In the uplink, the limiting factor for supporting a similarly high number of users is the noise rise. For such a high number of users in the cell it can be assumed that many users are not transmitting any user data for some time (e.g., for reading during web browsing or in between packets for periodic packet transmission such as VoIP). The corresponding overhead in the noise rise caused by maintained control channels will limit the number of users that can be efficiently supported. Since completely releasing the dedicated connection during periods of traffic inactivity would cause considerable delays for reestablishing data transmission and a correspondingly worse user perception, the Continuous Connectivity for Packet Data Users intends to reduce the impact of control channels on uplink noise rise while maintaining the connections and allowing a much faster reactivation for temporarily inactive users. This is intended to significantly increase the number of packet data users (i.e. HSPA users) in the UMTS FDD system that can stay in the active state (Cell_DCH) over a long time period, without degrading cell throughput. The objective aims also at improving the achievable UL capacity for VoIP users with its inherent periodic transmission through reducing the overhead of the control channels. 19
    19. 19. Figure 10. Relative uplink capacity cost of a user in the active state and not transmitting any data41 Thus far, the 3GPP study summarized in Fig. 10, indicates that with Rel-7, the uplink can support the users not transmitting any data in the active state with one third of the capacity cost than in Rel-6. These users see practically no delay when resuming the data transmission. The same studies indicate the potential for VoIP capacity increase in the uplink is in the order of 50% with Rel-7 (see Fig. 11). Figure 11. Relative uplink VoIP capacity with Rel-6 and Rel-742 The key technologies studied under the Continuous Connectivity for Packet Data Users for 3GPP Rel-7 are reducing the uplink noise rise caused by the uplink DPCCH transmission of the users not transmitting any data, by either reducing the uplink DPCCH transmission power (SIR_target reduction) or by shutting it down completely for short time durations (UL DPCCH gating). Capacity investigations for the latter are shown in the two figures above. The benefit of turning the UE’s transmitter off is significantly reduced UE power consumption which also enables longer times for the UE to be kept in the active state when the applications require no data transmission. The benefit of just reducing the uplink DPCCH power to a certain level is that synchronization and power control can be kept stable in all situations together with a less bursty behavior of the channel. In connection with the key technologies further enhancements (e.g., CQI reductions, UL DPCCH slot format modifications) are considered to improve the UL noise rise situation. 41 3GPP TR 25.903 v1.0.0 (2006-05), Continuous Connectivity for Packet Data Users 42 Ibid. 20
    20. 20. Although the main trigger of the work on "continuous connectivity for packet data users" was the bottleneck of UL noise rise which is especially costly for temporarily inactive users, further control channel overhead reduction improvements are also investigated for the downlink (e.g., when looking at UE battery savings or VoIP capacity enhancements). The downlink HSPA VoIP capacity enhancement is being investigated; the key techniques being optimizing the control signaling of HSDPA to be most efficient for small VoIP-like packets and optimizing the HSDPA scheduler to take the full advantage of knowing that certain traffic is VoIP traffic. To some extent, the scheduler enhancements and control signaling optimizations may be mutually exclusive and thus selecting which method is best requires careful consideration. Gaming over IP (GoIP) As part of the IMS realtime services using HSDPA/HSUPA, GoIP may become very important applications for the operator. To ensure the target gaming performance and propose proper improvements of existing standards, it is necessary to investigate and define the performance requirements for gaming traffic. Existing requirements The following tables are cited from the current 3GPP standards: Table 1. 3GPP TS 22.105: requirements for conversational voice, videophone and interactive games Medium Application Degree of Data rate Key performance parameters and target symmetry values End-to-end Delay Information One-way Variation loss Delay within a call Audio Conversational voice Two-way 4-25 kb/s < 150 msec < 1 msec < 3% FER preferred < 400 msec limit * Video Videophone Two-way 32-384 kb/s < 150 msec < 1% FER preferred < 400 msec limit Lip-synch < 100 msec Data Interactive Two-way < 1 KB < 250 msec N.A Zero games *The overall one way delay in the mobile network (from UE to PLMN border) is approximately 100 msec If VoIP over HSDPA is intended to become the primary mechanism for providing voice and/or video services, then it can be assumed that the requirements should not be degraded from the existing requirements shown above. Requirements for interactive games are obviously very dependent on the specific game, but it is clear that demanding applications will require very short delays. Consistent with demanding interactive applications, a maximum 250ms end-to-end (E2E) one-way delay is proposed here and leads to a ping time of 500ms. 21
    21. 21. Table 2. 3GPP TS 23.107: RAB characteristics for Conversational class RAB characteristics Traffic class Conversational class Maximum bitrate (kbps) <= 16 000 (2) (7) Delivery order Yes/No Maximum SDU size (octets) <=1 500 or 1 502 (4) SDU format information (1) (5) Delivery of erroneous SDUs Yes/No/- Residual BER 5*10-2, 10-2, 5*10-3, 10-3, 10-4, 10-5, 10-6 SDU error ratio 10-2, 7*10-3, 10-3, 10-4, 10-5 Transfer delay (ms) 80 – maximum value Guaranteed bit rate (kbps) <= 16 000 (2) (7) Traffic handling priority Allocation/Retention priority 1,2,3 (1) Source statistic descriptor Speech/unknown Signalling Indication No specific mention of gaming traffic has been made. The transfer delay characteristic is one way, and measured from CN to UE. This allows for 150ms for internet based delay which is more than feasible. Gaming over HSDPA The following three major performance parameters have been identified to be sufficient for defining the gaming performance requirements: • End-to-end delay • Jitter • Application packet loss Given the wide variety of games available, it is recommended to separate the games into the following four categories in which the corresponding tolerable gaming performance have been described with the use of the above proposed performance parameters: • First Person Shooter (FPS) – fast user response, many online players at once, highly dynamic - Up to 150ms end-to-end delay may be acceptable - 10ms jitter may be expected to be critical for FPS gaming - Up to 5% packet loss is acceptable • Real Time Strategy (RTS) – slightly lower response required, slower gameplay, handful of players in a single game - 250ms-500ms end-to-end delay may be acceptable - Requirements for jitter are undefined - 1% packet loss with 150ms delay may be acceptable • Massive Multiplayer Online Role Playing Games (MMORPG) – Persistent games, highly variable scenarios, many hundreds of players online at once, many tens in a given situation - Several packets every ms - Latency << 350ms o With 80ms RAB latency no problem o 150ms ping delay cannot be achieved with 80ms RAB - Depending on the time and game contents, the required data rate ranges between 8kbps- 24kbps 22
    22. 22. - 10% packet loss may be acceptable if latency is low; 10~15% packet loss with 150ms ping delay is acceptable • Non-real Time Games (NRTG) – e.g., chess, backgammon, cards etc. - Zero packet loss, which can be achieved by retransmission methods It is understood that, within these four gaming categories, the requirements for delay, jitter and packet loss are different depending on the games and on the expectations of the player, but the typical range of the requirement attributes for real time gaming have been observed as following: • Packet loss 0.1% 5% • Latency (e2e) 75ms – 250ms • Data rate (5kbps- 60kbps) It is clear from the values above that the definition in TS 22.105 is considerably stricter than necessary in terms of packet loss in most cases. Jitter has been shown to be problematic, but not in a truly quantifiable way and it is proposed that jitter should be minimized to as little as possible, potentially by algorithms in the application rather than in the network. However, in terms of latency, the requirements on the RAB delay for best performance are too low since E2E delays >75ms will lead to a perceived drop in performance from the user if the game requires the fastest response. Activities in 3GPP on defining more appropriate values related to realtime gaming for TS 22.105 are ongoing. Currently, the following requirements are in discussion: Table 3. 3GPP TS22.105 Values Related to Realtime Gaming Medium Application Degree of Data End-to-end Delay Information symmetry rate One-way Variation loss Delay within a call Data realtime two-way < 60 < 75 msec N.A. < 3% FER games kb/s preferred preferred < 5% FER limit It should be noted that these values are considered the most demanding with respect to delay requirements (e.g., supporting First Person Shooter games). Other types of games may require higher or lower data rates and more or less information loss but can tolerate longer end-to-end delay. Delay optimization for procedures applicable to PS and CS Connections In Rel-99, UMTS introduced a dedicated channel (DCH) that can be used for CS and PS connections when UE is in CELL_DCH state. In addition to CELL_DCH state, Rel-99 introduced CELL_FACH state where signaling and data transmission is possible on common channels (RACH and FACH) and CELL_PCH and URA_PCH states, where the transmission of signaling or user data is not possible but enables UE power savings during inactivity periods maintaining the RRC connection between UE and UTRAN and signaling connection between UE and PS CN. The introduction of the CELL_PCH and URA_PCH states, the need of releasing the RRC connection and moving the UE to Idle mode for PS connections was removed and thus the Rel-99 UTRAN can provide long living Iu-connection PS services. On the other hand, when UE is moved to CELL_PCH or URA_PCH state, the start of data transmission again after inactivity suffers inherent state transition delay before the data transmission can continue in CELL_DCH state. As new packet-oriented features like HSDPA and HSUPA in Rel-5 and Rel-6 UMTS systems respectively provide higher data rates for both downlink and uplink in CELL_DCH state, the state transition delay has been considered to be significant and negatively influencing the end user experience. In addition to RRC state transition delay, the radio bearer setup delay to activate new PS and CS services has been seen as problematic in UMTS, due to signaling delays on CELL_FACH state where only low data rates are available via RACH and FACH, and due to activation time used to synchronize the reconfiguration of the physical and transport channel in CELL_DCH state. 23
    23. 23. To secure future competitiveness of UMTS and enhance the end user experience even further, the delay optimization for procedures applicable to PS and CS connections work is targeted to reduce both setup times of new PS and CS services and state transition delays to, but still enable, excellent UE power saving provided by CELL_PCH and URA_PCH states. During the 3GPP Rel-6 time frame, the work was focused on solutions that can be introduced in a fast manner on top of existing specifications with limited effects to the existing implementations. In addition, the solutions which allow the Rel-6 features to be used in the most efficient manner were considered. The agreed modifications can be summarized as: introduction of enhanced support of default configurations, reduced effects of the activation time, and utilization of HSPA for signaling. Thus, from Rel-6 onwards, the signaling radio bearers (SRBs) can be mapped on HSDPA and HSUPA immediately in RRC connection setup and default configurations can be used in radio bearer setup message and RRC connection setup message in a more flexible manner. The utilization of default configuration and mapping of the SRBs on HSDPA and HSUPA will reduce message sizes, activation times, and introduce faster transmission channels for the signaling procedures, thereby providing significant enhancement to setup times of PS and CS services compared to Rel-99 performance. In the 3GPP Rel-7 time frame, the work will study methods of improving the performance even further, especially in the area of state transition delays. As the work for Rel-7 is less limited in scope of possible solutions, significant improvements to both RRC state transition delays and service setups times are expected. 4.3 IMS/Core Network Enhancements Fixed Broadband Access to IMS (FBI) It is recognized that several standards organizations are in the process of defining NGN (Next Generation Network) session control using IMS as a platform. Fixed Broadband Access provides a means of extending IMS services to wireline subscribers. The work on Fixed Broadband Access (FBI) will embed IMS as the framework for advanced services for many types of operators. IMS is extended to act as a common core and services platform for fixed and mobile networks. Under the Fixed Broadband Access (FBI) work, 3GPP is considering requirements from a number of different communities of operators including those based on DSL and cable technology. IMS is now accepted as the de facto basis for session oriented services in next generation telecommunications networks. A number of features have been introduced primarily in order to extend the IMS towards providing services for fixed broadband access: • NAT traversal • Optional use of SIP preconditions for wireline access • PSTN bridging – the use of IMS as a transit network with PSTN round the edge One Tunnel for Packet Traffic The amount of user plane data is assumed to increase significantly during the next few years because of the introduction of HSPA and the IP Multimedia Subsystem. Currently in UMTS, packet data traffic must traverse two nodes in the core network – the SGSN and GGSN. This is independent of the traversal done in the UTRAN. The discussions in 3GPP have shown that more scalable UMTS system architecture can be achieved by direct tunneling of user plane data between the RNC and the GGSN, which is known as One Tunnel approach (Fig. 12). In the One Tunnel Approach, the transport and control functionality of the SGSN are separated, resulting in a new distribution of functionality with the GGSN. The new SGSN controller (cSGSN) is performing all control functions of an SGSN while the enhanced GGSN (xGGSN) is responsible for SGSN and GGSN transport functionality for the majority of traffic. 24
    24. 24. cSGSN xGGSN Internet or UTRAN intranet Control plane User plane Figure 12. One Tunnel architecture This architecture has several benefits: • All bearer resources are shared between cSGSNs resulting in reduction in CAPEX by eliminating the bearer hardware on new cSGSNs • 3G bearer plane is engineered only at the xGGSN and not tied to geographic distribution of users • User plane scalability: as 3G bearer throughput grows, bearer resources only added at the xGGSN • No need to project 3G bearer capacity of individual cSGSN for roamers; only need to project network-wide 3G bearer capacity for roamers IMS Multimedia Telephony Communication Enabler (MITE) and Supplementary Services The aim of MITE is to provide specifications for common telephony services over IMS, including well known and well used supplementary services. This builds on the capabilities provided by IMS in Rel-5 and Rel-6. MITE will enable consistent user experience and interoperable implementations. MITE covers voice and multimedia user-to-user sessions that can be enhanced with features like forwarding and conferencing. MITE offers operators a path to evolve telephony services from the CS domain towards IMS without abandoning legacy features. MITE is aligned with service definitions being generated for Next Generation Networks (NGN) and therefore MITE can be used in conjunction with the FBI work item to provide consistent wireline and wireless services. Combining CS and IMS (CSI) CSI provides a means for a mobile station to combine a CS voice call with PS-based IMS services between the same two users. The motivation is to introduce multi-media telephony, without the immediate requirement to upgrade the radio to support VoIP. This is done by using legacy CS voice for the voice component of the multi-media session. In addition, CSI introduces a SIP-based mechanism for users to exchange terminal capabilities. Capability exchange provides a means by which party A can know which IMS services/media types can be included as part of a multi-media session with party B. Capability exchange can be done independent of CSI and is expected to be used for detecting if a voice call can be successfully upgraded to video. CSI provides a path to introduce subscribers to IMS services and in particular the concept of rich calls. The main usage scenario is to start each session with a CS voice call then to detect what add-on services are possible and display these available services to the user in order to encourage their use. By re-using CS calls for the voice, there is no risk of VoIP performance issues impacting IMS launch. CSI also allows operators to manage their investment in rolling out IMS while establishing the level of market demand for new services. CSI is the basis for video sharing services (adding video to an established voice call in progress) being launched on some networks. 25
    25. 25. Policy and Charging Convergence (PCC) PCC provides enhanced and future-proof tools to allow operators to perform advanced dynamic QoS control and charging for IP data bearers. PCC develops the capabilities provided in Rel-5 Service Based Local Policy (SBLP) and Rel-6 Flow Base Charging for IP bearers. PCC will combine the policy control and charging functions into a single architecture which provides consistent signaling and which can support multiple access types. PCC will enable operators to adjust the QoS being provided to users based on a range of factors including application requirements and user subscription level (gold, silver, bronze). PCC will also provide tools useful to manage the impact of “bandwidth hogging” applications such as peer-to-peer file sharing. The PCC architecture is built on the Rel-6 Flow Based Charging including evolution of the Rel-6 FBC reference points (Gx, Gy, Rx). Like FBC (Flow Based Charging), the PCC architecture can manage multiple service data flows on one single bearer, allowing more flexibility to configure and optimize the radio interface. Voice Call Continuity (VCC) VCC is a home IMS application, which enables seamless continuity of voice services between CS domain and IMS. VCC subscribers’ calls are anchored in home IMS when roaming across CS domain and IMS. VCC provides functions for call originations, terminations and call continuity across CS domain and IMS. The VCC Application is inserted in the control signalling path of VCC subscriber’s CS and IMS sessions for employment of a 3pcc (third party call control function) controlling these sessions. Calls established over CS domain are redirected to IMS for insertion of VCC Application in the call control signalling path using standard CS domain techniques available for redirecting calls at call establishment; these calls are then processed according to standard IMS procedures. This allows for voice service continuity across CS domain and IMS. VCC provides selection of domain for delivery of incoming calls to VCC subscribers simultaneously registered in CS domain and IMS. VCC provides inter-domain mobility while maintaining active session(s), manifested by transfers between CS domain and IMS, with capability to transition multiple times in both directions. 5 Evolution Beyond Rel-7 – SAE/HSPA+/LTE To ensure the competitiveness of the 3GPP systems in a time frame of the next ten years and beyond, a long term evolution of the 3GPP access technology is specified in 3GPP. The vision of this long term evolution and transition towards a packet switched only system is shown in Fig. 13. 2006 2008-9 2012-17 Current Phase Transition Phase PS-Only Phase - CS Domain continues to be - Network Architecture is progressively - Large Majority of the the primary enabler of migrating towards a Packet-based Traffic is carried over Conversational Services Architecture optimized to carry all QoS PS Domain - PS Domain restricted to Data Classes of Traffic (from Best Effort to - CS Domain is only used Applications and some Conversational) for small proportion of innovative RT/near-RT - Legacy CS Traffic is increasingly carried legacy Devices/Users Services over PS Domain - PS Transition completed Figure 13. Migration towards full service delivery through the evolved packet core43 To enhance the capability of the 3GPP system to cope with the rapid growth in IP data traffic, the packet- switched technology utilized within 3G mobile networks requires further enhancement. Continued evolution and optimization of the system is also necessary in order to maintain a competitive edge in terms of both performance and cost. Important parts of such a long term evolution include reduced 43 Cingular Wireless 26
    26. 26. latency, higher user data rates, improved system capacity and coverage, and reduced overall cost for the operator. Additionally, it is expected that IP-based 3GPP services will be provided through various access technologies. A mechanism to support seamless mobility between heterogeneous access networks is needed for future network evolution. In order to achieve this, an evolution or migration of the network architecture, as well as an evolution of the radio interface is studied in the System Architecture Evolution (SAE), Long Term Evolution (LTE) and HSPA Evolution (HSPA+) Study Items in 3GPP. From a network deployment point of view it is likely that HSPA enhancements will be introduced first, followed by the evolved packet core (SAE) and then the evolved radio interface (LTE). 5.1 Architecture Evolution (SAE) SAE or 3GPP System Architecture Evolution is the study item where the 3GPP industry is developing a framework for an evolution and migration of current systems to a higher-data-rate, lower-latency, packet- optimized system that supports multiple radio access technologies. The focus of this work is on the PS domain with the assumption that voice services are supported in this domain. The main drivers for the network evolution are: • To be able to meet the targets for the evolution of the radio-interface (a.k.a. LTE) • To enable the evolution towards an All-IP Network • To support mobility and service continuity between heterogeneous access networks The main objectives for the evolved architecture are: • To get very low latency for the overall network • To efficiently support a variety of services, especially from the PS domain (e.g., Voice over IP, Presence) • To support a variety of different access systems (existing and future) and support access selection based on combinations of operator policies, user preferences and access network conditions • To improve basic system performance e.g., communication delay, communication quality, connection set-up time, etc. • To be able to maintain negotiated QoS across the whole system; in particular inter-domain and inter-network interworking and QoS on the network link to the Base Station site • To support service continuity between I-WLAN and the 3GPP PS domain • To support multiple radio access technologies and terminal mobility between different radio access technologies • To be able to maintain and support the same capabilities of access control (authentication, authorization), privacy and charging when moving between different radio access technologies High Level Architecture Overview Figure 14 shows the current architecture model for the evolved system. It identifies functional entities which may be physically co-located or distributed, according to product development and deployment scenarios. For instance, the Inter-AS anchor may or may not be co-located with the UPE; the MME and UPE may or may not be co-located in the same node. 27
    27. 27. GERAN Gb SGSN GPRS Core PCRF Iu UTRAN Rx+ S7 S3 S4 HSS Op. S5a S5b S6 IP S1 MME 3GPP SAE SGi Serv. Evolved RAN Anchor Anchor (IMS, UPE IASA PSS, Evolved Packet Core etc…) S2 S2 non 3GPP WLAN IP Access 3GPP IP Access * Color coding: red indicates new functional element / interface Figure 14: Logical high level architecture for the evolved system44 New functional elements The architecture model contains the following new functional elements: • Mobility Management Entity (MME): The MME manages and stores UE context (for idle state: UE/user identities, UE mobility state, user security parameters). It generates temporary identities and allocates them to UEs. It checks the authorization whether the UE may camp on the TA or on the PLMN. It also authenticates the user. • User Plane Entity (UPE): The UPE terminates for idle state UEs the downlink data path and triggers/initiates paging when downlink data arrive for the UE. It manages and stores UE contexts, e.g., parameters of the IP bearer service or network internal routing information. It performs replication of the user traffic in case of interception. • 3GPP Anchor: The 3GPP Anchor is a functional entity that anchors the user plane for mobility between the 2G/3G access system and the LTE access system. • SAE Anchor: The SAE Anchor is a functional entity that anchors the user plane for mobility between 3GPP access systems and non-3GPP access systems. New interfaces The architecture model contains the following new interfaces: • S1: S1 provides access to Evolved RAN radio resources for the transport of user plane and control plane traffic. • S2: S2 provides the user plane with related control and mobility support between WLAN 3GPP IP access or non 3GPP IP access and the SAE Anchor. • S3: S3 enables user and bearer information exchange for inter 3GPP access system mobility in idle and/or active state. It is based on Gn reference point as defined between SGSNs. • S4: S4 provides the user plane with related control and mobility support between GPRS Core and the 3GPP Anchor and is based on Gn reference point as defined between SGSN and GGSN. • S5a: S5a provides the user plane with related control and mobility support between MME/UPE and the 3GPP Anchor. • S5b: S5b provides the user plane with related control and mobility support between the 3GPP anchor and SAE anchor. 44 3GPP TR 23.882 v1.2.0 28